Impact of the UK QE on banks’ balance sheets – Bank Underground

Mahmoud Fatouh

Quantitative easing (QE) entails creating new central financial institution reserves to fund asset purchases. Deposited within the reserves account of the vendor’s financial institution, these reserves can have implications for banks’ asset mixes. In our paper, we use stability sheet information for 118 UK banks to empirically examine whether or not the asset compositions of banks concerned within the UK QE operations reacted in a different way compared to banks not concerned within the preliminary rounds of QE between March 2009 and July 2012.

QE and central financial institution reserves

QE consists of the creation of central financial institution reserves (reserves hereafter), which significantly will increase the dimensions of central financial institution stability sheet. Chart 1 reveals that reserves within the UK elevated considerably after the introduction of UK QE.

Chart 1: Financial institution of England liabilities and capital (£million, weekly)

Supply: Financial institution of England

When the Financial institution of England conducts QE, reserves are credited to the reserves account of the vendor’s financial institution, and that financial institution then credit the vendor’s deposit account with the identical quantity. Therefore, banks concerned in QE operations (QE banks) initially obtain further liquidity (as reserves), whereas different banks (non-QE banks) don’t.

Nevertheless, the distribution of the extra reserves between QE banks and non-QE banks might change. A number of the further liquidity might ‘leak’ to non-QE banks because the sellers (principally non-bank financials within the UK) make investments their cash in different property, ensuing of their deposits shifting across the banking system. The magnitude of this ‘leak’ relies on whether or not these deposits find yourself with different QE banks or not

In the end, we argue that whereas a few of the further reserves would leak to non-QE banks, most would stick with QE banks. It is because non-bank financials principally do enterprise with a small subset of banks who’re additionally members within the Financial institution’s QE operations. As a result of dynamic nature of reserves and the a number of components affecting them, it’s arduous to isolate the influence of QE on the distribution of further reserves between QE banks and non-QE banks. But, evaluating the inventory of reserves earlier than and after QE introduction helps our argument (see Chart 2).

Chart 2: Modifications in lending and securities – QE banks versus non-QE banks

QE and financial institution lending

Conventionally, within the presence of reserve requirements, any improve in reserves can probably improve complete credit score supplied by the banking system by a number of multiples of the preliminary improve in reserves (money multiplier effects). This could be the case if there are not any different constraining components corresponding to capital and liquidity necessities (which are extra binding than reserve necessities), or low demand for credit score. There are not any reserve necessities within the UK (ie the cash multiplier is undefined), that means that the availability of credit score is principally pushed by banks’ capability and/or incentives to lend.

Whereas the supply of liquidity and capital positions govern banks’ capability to lend, their incentives are largely decided by the regulatory framework in operation particularly in downturns. Put up-crisis, the extra reserves from QE purchases elevated the supply of liquidity which alleviated liquidity constraints. Nevertheless, banks had depleted capital positions and operated in a regulatory framework that assigns low danger weights to funding in authorities securities and far larger weights to lending to the true financial system. Consequently, elevated reserves could or could not have led to larger lending, and therefore the Financial Coverage Committee didn’t emphasise the transmission of QE impact via the bank lending channel (BLC).

In different phrases, QE improves banks’ capability to lend when carried out in a liquidity-scarce setting, however different components, corresponding to financial institution capital positions and danger weighting, can have an effect on the influence of QE on financial institution lending.

What we do

We’re taken with understanding the influence of QE, whereas controlling for different developments that would additionally have an effect on financial institution lending, such because the post-crisis Basel III reforms. These reforms tightened capital necessities, requiring weakly capitalised banks to boost further capital assets. To isolate the influence of QE, we assemble a difference in differences (DiD) strategy, through which we evaluate modifications in stability sheets of QE banks (handled group) to these of comparable non-QE banks (management group). We decide QE banks utilizing a confidential Financial institution of England’s information set, which reveals which banks obtained reserves by way of UK-QE operations and the dimensions of the extra reserves. Our pattern consists of stability sheet information for 118 UK banks from 2000 to 2018.

To attract significant conclusions concerning the position performed by QE, we have to assess the (statistical) significance of the variations between the 2 teams, isolate the influence of different components, and guarantee any results we observe are particular to the UK-QE interval.

It is very important observe that our evaluation covers the influence of the early QE rounds, ie these till July 2012. The consequences of the Brexit and Covid-19 rounds are coated in one other paper assessing the interplay between QE and the federal government lending help schemes throughout the Covid stress.

QE banks versus non-QE banks

QE banks are on common greater and maintain comparatively extra securities than non-QE banks. Therefore, with none modifications, our outcomes could be susceptible to choice bias, within the sense that any variations detected between QE banks and non-QE banks might be resulting from variations in financial institution traits fairly than QE therapy. To alleviate the results of choice bias on the validity of outcomes, we use a propensity score matching methodology to remove common variations between the therapy and management teams, and therefore enhance the validity of our conclusions primarily based on comparisons between the 2 teams. Propensity scores depend on completely different financial institution traits, and are used to create a matched set from non-QE banks for every QE financial institution, primarily based on a sure matching ratio (1:5 in our baseline setup). That’s, every QE financial institution is assigned a matched set that features 5 non-QE banks with propensity scores most just like it. Matching would achieve success if it manages to remove pre-matching variations between the therapy and management teams. We examine that by regressing a therapy standing dummy on variables reflecting the dimensions, profitability and stability sheet profile, earlier than and after matching. As Desk A illustrates, with out matching, QE banks, on common, are bigger and maintain extra securities, and these variations are statistically vital. Put up matching, common variations between QE banks and the (matched) management group develop into statistically insignificant. We run our matching utilizing different matching ratios from 1:1 to 1:8, and observe related patterns.

Desk A: Propensity rating matching

Notes: Probit regressing the therapy on financial institution traits in 2008h2. The dependent variable is the financial institution therapy standing. The unbiased variables are measurement because the pure log of complete property, fairness as complete property minus complete liabilities, return on property (ROA), complete securities over complete property and web curiosity earnings over complete property. Mannequin (1) studies the pre-matching outcomes whereas mannequin (2) studies the submit matching outcomes with matching ratio 1:5. Coefficients and normal errors are reported for every variable. Commonplace errors are clustered on the financial institution degree and reported in brackets, * p<0.10 ** p<0.05 *** p<0.01. 

DiD Outcomes

Our DiD mannequin controls for variations in measurement (complete property), degree of leverage (fairness to complete property), profitability (return on property), and securities to complete property and web curiosity earnings to complete property (variations in enterprise fashions).

Financial institution lending

Desk B presents the therapy coefficients for lending DiD regressions. According to the common tendencies in Chart 2, therapy coefficients for all (however one) lending regressions are statistically insignificant. That’s, we discover no proof of the choice financial institution lending channel (BLC); the extra liquidity didn’t incentivise QE-bank to extend lending, relative to the management group. There is no such thing as a proof suggesting that these outcomes had been pushed by modifications in relative demand for loans the 2 teams confronted.

Desk B: Therapy coefficients for DiD lending regressions

Notes: Therapy standing (Handled_i) equals to 1 for QE banks and 0 for non-QE banks. Controls are measurement as log of complete property, fairness over complete property, return on property (ROA), securities over complete property and web curiosity earnings over complete property. The reported p-values check the coefficient inequality between QE1 and QE2. Commonplace errors are clustered on the financial institution degree and reported in brackets, * p<0.10 ** p<0.05 *** p<0.01.

Securities and different property

Desk C reveals the DiD coefficients for different financial institution property. Relative to the management group, QE banks elevated reserves and lowered lending to different banks after QE1. Additionally they elevated holdings of presidency securities, particularly after QE2. This implies that QE banks reallocated their assets from lending in direction of authorities securities with low danger weights.

Desk C: Therapy coefficients for DiD regressions for different property

Notes: Therapy standing (Handled_i) equals to 1 for QE banks and 0 for non-QE banks. Controls are measurement as log of complete property, fairness over complete property, return on property (ROA), securities over complete property and web curiosity earnings over complete property. The reported p-values check the coefficient inequality between QE1 and QE2. Commonplace errors are clustered on the financial institution degree and reported in brackets, * p<0.10 ** p<0.05 *** p<0.01.

Conclusion

We check whether or not the extra reserves created through QE led to a rise in financial institution credit score. We accomplish that by evaluating the behaviour of QE banks’ property, in comparison with different banks, utilizing a DiD strategy.

We discover no proof of transmission through the BLC. We expect that the optimisation of regulatory capital motivated QE banks to speculate the extra liquidity in high-yield low danger weight sovereigns fairly than enterprise loans that appeal to larger danger weights.

Mahmoud Fatouh works within the Financial institution’s Prudential Framework division.

If you wish to get in contact, please electronic mail us at [email protected] or go away a remark under.

Comments will solely seem as soon as authorized by a moderator, and are solely revealed the place a full title is provided. Financial institution Underground is a weblog for Financial institution of England employees to share views that problem – or help – prevailing coverage orthodoxies. The views expressed listed here are these of the authors, and aren’t essentially these of the Financial institution of England, or its coverage committees.